New research shows that the rapid expansion of hyperscaler datacenters in the US has led to significant carbon emissions. These datacenters, primarily used for artificial intelligence, emitted about 25.7 million metric tons of carbon dioxide in a year, equivalent to Denmark's annual emissions. The study highlights that these facilities consume a large amount of electricity, with over half coming from fossil fuels.
The study emphasizes that the emissions from these datacenters are likely underestimated, as the number of facilities has continued to grow. This growth has raised concerns among climate groups, as the reliance on fossil fuels contributes to global heating. The rapid expansion of these datacenters has also sparked local opposition due to concerns about noise, air pollution, and water usage.
The Stratos datacenter in Utah is a notable example, with potential health-related economic costs of up to $690 million annually due to air pollution. The facility's developers are considering using combined-cycle natural gas but are also exploring renewable options. The study calls for further research and policy considerations to address the environmental impact of these datacenters.
This summary was written with AI assistance from the original article and may leave out important context. The original reporting belongs to The Guardian · Health. Check the original, and official sources, before acting on any detail.